calculatePredictors: Calculate predictors from MSG cloud masked data

In environmentalinformatics-marburg/Rainfall: Functions for training, predictiion and validation of an MSG based optical rainfall retrieval

Description

Calculate predictors from MSG cloud masked data

Usage

calculatePredictors(scenerasters, model = NULL, useOptimal = TRUE,
  spectral = NULL, sunzenith = NULL, texture = NULL, pptext = NULL,
  zonstat = NULL, filterstat = NULL, shape = NULL, min_x = NULL,
  max_x = NULL, further = c("sunzenith", "jday"), date)

Arguments

scenerasters	A raster stack of cloud masked MSG scenes with non clouded areas were set to NA
model	A rfe object Optional. Can be used instad of the spectral,texture, pptext,zonstat,shape parameters. Variables included in the optimal model are the calculated.
useOptimal	if is.null(model): Logical. Use the optimal variables from rfe or those less variables which lead to a model performance within one sd of the optimal model?
spectral	A character vector indicating the msg channels to be included. Possible values: "VIS0.6","VIS0.8","NIR1.6","IR3.9","WV6.2","WV7.3","IR8.7", "IR9.7","IR10.8","IR12.0","IR13.4","T0.6_1.6","T6.2_10.8","T7.3_12.0", "T8.7_10.8","T10.8_12.0", "T3.9_7.3","T3.9_10.8"
sunzenith	A raster of the sun zenith values
texture	data frame of all spectral and texture combinations ("mean", "variance", "homogeneity", "contrast", "dissimilarity", "entropy","second_moment") and filter sizes which are to be calculated. (Tip: Use expand.grid to create this data.frame)
pptext	data frame of all spectral and texture combinations which are to be calculated for teh overall cloud entity. (Tip: Use expand.grid to create this data.frame)
zonstat	data frame of all spectral and zonal stat ("mean","min","max","sd") combinations (min,max,mean or sd) which are to be calculated for the overall cloud entity. (Tip: Use expand.grid to create this data.frame)
filterstat	data.frame of all spectral and "mean","min","max","sd" and filter size combinations
shape	geoemtry variables which should be included. Possible values are "Ar",SI","CA","Ur","CAI","PAR","distEdges","Re","Ru","OIC", CI1","CO1","CI2","CO2","CCI1","CCI2","CO","SHD","C1","E", "TR","CR","C2","FR","EI","SF1","GSI","SF2","C3","SF3"
further	a character vector including Currently "jday" and/or "sunzenith" which will also be used as variables. see geometryVariables and borgIndices for description of the variables.
date	Date of the msg scene in format yyyymmddhhmm. Only imprtant if the day of the year (jday) is calculated (see param "further").
x_min	see textureVariables
x_max	see textureVariables

Author(s)

Hanna Meyer

Examples

############################################################################
#Example 1: Predictors from predictor list
############################################################################ 
# stack the msg scenes:
msg_example <-getChannels(inpath=system.file("extdata/msg",package="Rainfall"))

# raster the sunzenith 
sunzenith<-getSunzenith(inpath=system.file("extdata/msg",package="Rainfall"))

#get Date
date <- getDate(system.file("extdata/msg",package="Rainfall"))

#calculate variables (takes some time...)
pred <- calculatePredictors(msg_example,
sunzenith=sunzenith,
spectral=c("VIS0.6","VIS0.8","NIR1.6","T0.6_1.6","T6.2_10.8"),
texture=expand.grid(c("NIR1.6","T6.2_10.8"),
c("variance", "contrast"),c(3,9)),
pptext=expand.grid("T3.9_10.8",c("variance","mean")),
shape=c("Ar","CAI","SI","CI1"),
filterstat=expand.grid(c("VIS0.6","T6.2_10.8"),
c("min", "max"),c(3,9)),
zonstat=data.frame("spec"=c("VIS0.8","VIS0.8","T6.2_10.8"),
"var"=c("min","sd","max")),
date=date)
print(pred)
############################################################################
#Example 2:calculate predictors from an rfe model
############################################################################
#'  # stack the msg scenes:
msg_example <-getChannels(inpath=system.file("extdata/msg",package="Rainfall"))

data(rfeModel)
pred<-calculatePredictors(msg_example,model=rfeModel,date=NULL,sunzenith=NULL)

environmentalinformatics-marburg/Rainfall documentation built on May 16, 2019, 7:49 a.m.